|
- 积分
- 0
- 威望
- 0
- 包包
- 0
|
作者：Matthias Kroiss, Marina Leyerer, Valentin Gorboulev, Thomas Kühlkamp, Helmut Kipp, and Hermann Koepsell作者单位：Institut für Anatomie und Zellbiologie, Bayerische Julius-Maximilians-Universität, Würzburg, Germany
5 v$ i& X! B% s5 ` ' w. h8 W2 M; `! ^- [
- }6 e4 p; n) k( e4 V. o9 F5 D8 K
) r" k/ J# A/ F1 E0 B ( L- j9 ~4 Q0 }. o5 A
' V/ r* T* l9 r6 |$ d/ F. @0 v - {+ x9 M- O$ s' M; D0 ^
$ r% T- T0 D; s# \8 q q! [
/ B3 _0 Q- k0 }
, ~9 H/ Z' r8 Q1 { S4 C
* M/ X- x6 G& Y2 X . n% t7 k, U# s# ]' h e# p G$ b% I
2 b7 Z0 p9 T! S5 H; O9 } 【摘要】
0 |- E8 p" j4 [3 `+ c# _8 a The product of gene RSC1A1, named RS1, is involved in transcriptional and posttranscriptional regulation of sodium- D -glucose cotransporter SGLT1, and removal of RS1 in mice led to an increase of SGLT1 expression in small intestine and to obesity (Osswald C, Baumgarten K, Stümpel F, Gorboulev V, Akimjanova M, Knobeloch K-P, Horak I, Kluge R, Joost H-G, and Koepsell H. Mol Cell Biol 25: 78-87, 2005). Previous data showed that RS1 inhibits transcription of SGLT1 in LLC-PK 1 cells derived from porcine kidney. A decrease of the intracellular amount of RS1 protein was observed during cell confluence, which was paralleled by transcriptional upregulation of SGLT1. In the present study, the subcellular distributions of endogenously expressed RS1 and SGLT1 were compared in LLC-PK 1 cells and human embryonic kidney (HEK)-293 cells using immunofluorescence microscopy. RS1 was located at the plasma membrane, at the entire trans -Golgi network (TGN), and within the nucleus. Treatment of LLC-PK 1 cells with brefeldin A induced rapid release of RS1 from the TGN, and confluence of LLC-PK 1 cells was accompanied by reduction of nuclear location of RS1; 84-90% of subconfluent cells and 5-34% of confluent cells contained RS1 in the nuclei. This suggests that confluence-dependent transcriptional inhibition by RS1 is partially regulated by nuclear migration. Furthermore, we assigned SGLT1 to microtubule-associated tubulovesicular structures and dynamin-containing parts of the TGN. The data indicate that RS1 inhibits the dynamin-dependent release of SGLT1-containing vesicles from the TGN.
3 ?% l3 o6 Q3 L; j7 f6 L 【关键词】 Na D glucose cotransport SGLT RS confluencedependent regulation nuclear migration dynamin brefeldin A5 x! Z3 W- Y1 _7 S: U% P
PREVIOUSLY, WE CLONED 67- to 68-kDa proteins from humans, pigs, rabbits, and mice that were termed RS1 (human gene RSC1A1 ) ( 17, 21, 23, 34 ). RS1 has a broad tissue distribution, including renal proximal tubules and small intestinal epithelial cells ( 17, 22, 23, 34 ). Western blots performed with subcellular fractions of Xenopus laevis oocytes in which RS1 was overexpressed suggested that RS1 is localized intracellularly and associated with the plasma membrane ( 16, 32 ).
6 l' @2 o8 T) |4 i& ?5 z- z+ c+ c2 A2 h1 [" ^
Measurements in porcine LLC-PK 1 cells provided evidence that RS1 is involved in transcriptional downregulation of sodium- D -glucose cotransporter SGLT1 in the subconfluent state ( 16, 28 ). Transcriptional upregulation of SGLT1 after reaching confluence was associated with a posttranscriptional decrease of the amount of RS1 protein. On reduction of RS1 via an antisense strategy, the expression of SGLT1 was upregulated, and, inversely, by overexpression of RS1, the expression of SGLT1 was largely decreased. In addition, nuclear run-off assays showed that the transcription of SGLT1 was increased in RS1-depleted cells ( 16 ). Experiments with oocytes of X. laevis revealed that RS1 is also a posttranscriptional inhibitor of the expression of SGLT1 and of some other plasma membrane transporters ( 17, 23, 35 ). The inhibition of human SGLT1 (hSGLT1) expression by hRS1 was abolished when a dominant-negative dynamin mutant was coexpressed ( 35 ). Given that dynamin is required for endocytosis and for vesicle budding from intracellular compartments such as endosomes or the trans -Golgi network (TGN) ( 12, 33 ), we could not distinguish whether the observed effects of RS1 are due to stimulation of dynamin-dependent endocytosis or to inhibition of cycling from an intracellular compartment. The selectivity of RS1 is not understood. However, evidence was obtained that SGLT1 is a physiologically important target of RS1. After removal of the Rsc1A1 gene encoding RS1 in mice, upregulation of SGLT1 and of glucose absorption in small intestine was observed, and the animals developed an obese phenotype ( 21 )., Y: J7 W8 }2 w: Z
6 K9 Q1 e4 F' Z8 Z" ?" z) b; `
In the present study, we compared the subcellular distributions of endogenously expressed SGLT1 and RS1 in two cell lines. LLC-PK 1 cells and human embryonic kidney (HEK)-293 cells were examined in both the subconfluent and confluent state. SGLT1 was located in the plasma membrane, in endosomal membranes, and in tubulovesicular membranes that are part of the TGN. At the TGN, RS1 was colocalized with dynamin. At variance, RS1 was localized at the intracellular side of the plasma membrane, at the entire TGN, and within the nuclei of subconfluent cells. Because dissociation of RS1 from the TGN was observed after treatment with brefeldin A (BFA), RS1 is likely to coat the TGN. Thus the plasma membrane and the TGN are compartments where interactions of RS1 and SGLT1 can occur. Together with previously described and recently observed functional characteristics of RS1, the present data strongly suggest that RS1 inhibits the delivery of SGLT1-containing vesicles from the TGN to the plasma membrane. They correlate nuclear migration of RS1 with transcriptional inhibition of SGLT1 in subconfluent cells, which is relieved after confluence.
3 i5 z; S7 y% b; {: `% e
" J% ?/ i6 T. u3 s+ j+ uMATERIALS AND METHODS% ?/ i4 R# _& Z
1 m6 m. q+ G E. R: O
Materials. Restriction enzymes were obtained from New England Biolabs (Frankfurt, Germany) and from MBI Fermentas (St. Leon-Rot, Germany). Triton X-114, benzamidine, aprotinin, leupeptin, phenylmethylsulfonyl fluoride, Igepal CA-630, colchicine, and proteasome inhibitor MG132 were supplied by Sigma (Taufkirchen, Germany). BFA was obtained from Calbiochem (Schwalbach, Germany) and 4',6'-diamidino-2-phenylindole (DAPI) from Molecular Probes (Leiden, The Netherlands). All other chemicals were supplied as described earlier ( 16 ).9 z& R' R6 \+ f, p& H5 v' T* C
9 U$ z/ y+ d) _9 o; X
Antibodies. Polyclonal antibody against recombinant porcine RS1 (pRS1-ab) was raised in rabbit as described previously ( 16, 32 ). In addition, a polyclonal antibody was also raised in rabbit against a peptide of pRS1 (pRS1-p-ab). This peptide (DKENVPRSRESVNESSC) is identical to amino acids 462-477 of pRS1 and contains an additional COOH-terminal cysteine. Affinity chromatography was performed against recombinant pRS1 protein (pRS1-ab) or against the antigenic peptide (pRS1-p-ab), both linked to Sulfolink-beads from Pierce (Rockford, IL). Polyclonal antibody against amino acids 243-272 of rabbit SGLT1 (QIS30) was raised in rabbit and affinity purified on antigenic peptide as has been described ( 14 ). Polyclonal antibody against green fluorescent protein (GFP) was obtained from Clontech Laboratories (Heidelberg, Germany), mouse monoclonal antibody against -tubulin of rat brain (TUB 2.1) was from Sigma, sheep anti-human TGN46 antibody was from Diagnostic International (Schriesheim, Germany), mouse monoclonal antibody against bovine brain clathrin (MAB1011) was from Chemicon International (Hofheim, Germany), and goat anti-human dynamin II was from Santa Cruz Biotechnology (Heidelberg, Germany). Goat anti-rabbit IgG F(ab') 2 coupled to AlexaFluor-555 and donkey anti-goat IgG F(ab') 2 coupled to AlexaFluor-488 were purchased from Molecular Probes; goat anti-mouse IgG coupled to Cy2 and donkey anti-sheep IgG coupled to Cy2 were obtained from Dianova (Hamburg, Germany).1 _ o b! ?7 V
3 n$ J; i, w" t) @DNA expression vectors and stably transfected LLC-PK 1 cells. For expression of GFP, the pEGFP-C1 vector was purchased from Clontech. To prepare GFP-pRS1 fusion protein with the NH 2 terminus of pRS1 linked to the COOH terminus of GFP, pBluescript II SK-pRS1 plasmid ( 16 ) was digested with Sac I and Apa I. The obtained 2.13-kb fragment of pRS1 cDNA was isolated and cloned into the pEGFP-C1 vector (pEGFP-C1-pRS1). This fragment contained the complete coding region and a 3'-noncoding part of pRS1. For stable transfection of LLC-PK 1 cells with pEGFP-C1 or pEGFP-C1-pRS1, subconfluent cells were transfected using lipofectin reagent from Gibco-BRL (Karlsruhe, Germany). After 2 days, transformants were selected by adding geneticin (G418) to the culture medium (0.4 mg/ml for 1 wk and 0.8 mg/ml later on).
1 ^" w' f' ~0 e' @/ f( _
, K! a" A5 @% B: \3 q: z/ O5 fSubcellular fractionation of LLC-PK 1 cells. For subcellular fractionation, LLC-PK 1 cells were washed with PBS (4°C) and scraped from the culture plates. The cells were suspended for 20 min in two volumes of ice-cold hypotonic HEPES buffer (H-buffer: 20 mM HEPES-HCl, pH 7.5, 10 mM potassium acetate, 15 mM magnesium acetate, 1 mM dithiothreitol, 1 mM benzamidine, 5 µg/ml aprotinin, 5 µg/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride), frozen in liquid nitrogen, and thawed three times. Finally the cells were homogenized in a glass Dounce homogenizer. The nuclei were sedimented from the homogenate by 5 min of centrifugation at 1,200 g, incubated for 5 min in ice-cold H-buffer supplemented with 0.5% (wt/vol) Igepal CA-630, and finally washed with H-buffer supplemented with 0.1% (wt/vol) Igepal CA-630. The 1,200 g supernatant of the homogenate was centrifuged at 40,000 g (1 h, 4°C), and a plasma membrane-enriched (PME) fraction was isolated as pellet. Associated proteins were removed by washing the pellet three times in H-buffer. Cytosolic proteins (Cy) were obtained in the supernatant after 2 h (4°C) of centrifugation of the 40,000 g supernatant at 200,000 g.2 D, K/ O. v7 D/ V! ^
# L5 {* C3 J/ U' _& ?( Z4 hCell culture. The HEK cell line 293 and the porcine renal epithelial (LLC-PK 1 ) cell line were maintained in DMEM supplemented with 10% (vol/vol) fetal bovine serum, 4 mM L -glutamine, 0.1 mg/ml streptomycin sulfate, and 100 U/ml penicillin G. Cells were grown at 37°C in the presence of 5% (vol/vol) CO 2 on petri dishes or on coverslips that were glued into the wells of six-well plates. The culture medium was replaced every 2-3 days. For passage, the HEK-293 cells were detached mechanically, whereas the LLC-PK 1 cells were detached by incubation for 10 min at 37°C in Ca 2 - and Mg 2 -free Dulbecco?s phosphate-buffered saline (DPBS; Sigma) supplemented with 28 mM NaHCO 3, 0.5 mM EDTA, and 10 mM HEPES, pH 7.4. The aspirated cells were pelleted by 5 min of centrifugation at 250 g and resuspended in culture medium. Strict monitoring of potential mycoplasmic contamination was carried out using the VenorGeM-PCR kit (Minerva Biolabs, Berlin, Germany).
2 b* O2 R! o* j& C" m& P
' z0 {* Y8 ^/ M: p% U0 ~Immunostaining. For immunostaining, HEK-293 cells and LLC-PK 1 cells were grown on coverslips to 50% confluence. The cells were washed twice with washing buffer [5 mM 3-( N -morpholino)propanesulfonic acid-NaOH, pH 7.4, 100 mM NaCl, 3 mM KCl, 2 mM CaCl 2, and 1 mM MgCl 2 ], fixed for 12 min with 4% (wt/vol) paraformaldehyde diluted in washing buffer, and washed twice again. Free aldehyde groups were quenched by 10-min incubation with washing buffer containing 40 mM glycine (pH 9). For immunoreactions, washed cells were permeabilized by a 10-min incubation with washing buffer containing 0.25% (wt/vol) Triton X-114 and incubated overnight at 4°C with primary antibodies diluted in washing buffer. The dilutions of primary antibodies were as follows: rabbit-anti-RS1-Ab, 1:50; QIS30 directed against SGLT1, 1:400; sheep-anti-TGN46, 1:125; goat anti-dynamin II, 1:50; and mouse-anti-clathrin, 1:200. After incubation with primary antibodies, cells were washed three times with washing buffer and incubated for 1 h at room temperature with fluorochrome-linked secondary antibodies, diluted as recommended by the suppliers. The cells were washed six times with washing buffer, rinsed shortly with double-distilled water, and embedded in Fluorescent Mounting Medium from DAKO Diagnostika (Hamburg, Germany) containing 1 µl of DAPI (Molecular Probes) per specimen for staining of the nuclei.
% a# |! G& N! o( @0 Z' X. Z4 W* V/ k; ^$ [+ l- Z, w4 G
The specificity of the antibodies was controlled as follows. The immunoreaction with affinity-purified pRS1-ab was abolished after preabsorption with the antigen by incubation of pRS1-ab for 60 min at 37°C with 0.1 mg/ml recombinant pRS1 protein. No antibody reaction with secondary antibodies was observed when the incubation with primary antibodies was omitted. In controls, no cross-reactivity of the secondary antibodies used with false primary antibodies used in the same experiment was detected.6 s/ o( p5 V6 {* _
0 m! n& Z0 r9 H- d& N0 o
Treatment of cells with BFA, colchicine, or MG132. To observe the effect of BFA on the intracellular distribution of RS1 and SGLT1, subconfluent LLC-PK 1 cells were incubated for 1 min, 5 min, or longer periods of time with culture medium containing 2 µg/ml BFA (stock solution of 2 mg/ml in methanol). Tubulin microfilaments were destroyed by incubation of subconfluent LLC-PK 1 cells for various time periods with 10 µM colchicine (stock solution of 10 mM in ethanol), whereas proteasomal degradation was inhibited by incubation of subconfluent HEK-293 cells grown on plates or confluent LLC-PK 1 cells grown on filters for 16 h with 10 µM proteasomal inhibitor MG132 (stock solution of 10 mM in DMSO). After incubation with BFA, colchicine or MG132 vesicular traffic was stopped by transfer of cells on ice and superfusion with 4°C cold washing buffer. Control cells were incubated with the respective concentrations of ethanol and methanol.
, O0 b6 i$ }9 f% b# M+ ]3 b( }$ q. b( R% w
Microscopy. For conventional fluorescence microscopy, we made use of an Axiophot-2 microscope (Zeiss, Jena, Germany) with standard Zeiss objectives and filter sets that was equipped with a SPOT RT charge-coupled device camera from Diagnostic Instruments (Visitron, Puchheim, Germany). The MetaView 6.2r4 software package from Universal Imaging (Visitron, Puchheim, Germany) was used for image acquisition, digital overlay, contrast enhancement, and measurements. A confocal laser-scanning imaging system, LSM 510, connected to an inverted microscope Axiovert-100 M equipped with the oil immersion objectives Plan Neofluar 40 x and Plan Apochromat 63 x (all from Zeiss, Jena, Germany) was employed. The lasers used (excitation wavelength and emission filter settings are given in parentheses) were as follows: UV laser for blue fluorochromes (364 nm, BP385-470 nm) from Coherent Enterprise (Santa Clara, CA), argon laser for green fluorochromes (488 nm, LP505 nm) from LASOS (Jena, Germany), and helium-neon laser (543 nm, BP560-615 nm) from LASOS. The system was controlled, and images were acquired and pseudocolored with the Zeiss LSM-510 software 2.5 SP2. Colors were chosen as blue for DAPI, red for AlexaFluor-555, and green for Cy2 and AlexaFluor-488; measurements were performed with the integrated tool.
( S' _0 \; A# `) {6 [' o/ q9 D5 F% e3 m! H4 t( K4 m( x6 q: Z
SDS-PAGE and Western blotting. Protein concentration was determined according to Bradford ( 2 ), using bovine serum albumin as a standard, and SDS-PAGE and Western blotting were performed as described earlier ( 16 ). For SDS-PAGE, protein samples were incubated for 5 min at 95°C in 50 mM Na 2 HPO 4, pH 6.8, 0.25 M -mercaptoethanol, 1% (wt/vol) SDS, and 0.0005% (wt/vol) bromophenol blue. For Western blots, affinity-purified primary antibodies were diluted 1:2,500 (pRS1-ab), 1:1,000 (pRS1-p-ab), 1:4,000 (QIS30), or 1:1,000 (antibody against GFP from Clontech). As secondary antibody, peroxidase-conjugated goat anti-rabbit IgG (whole molecule) antiserum from Sigma (1:5,000) was used. Western blotting, immunodetection using enhanced chemiluminiscence, and densitometric quantification of immunostained proteins in Western blots were performed as described ( 16, 32 ).4 w5 Q# @- p0 Z5 R
% r9 N" ]( M% N6 [* F
RESULTS" l! X1 q# s- Z
- G" E9 i7 ]* JSubcellular distribution of GFP-pRS1 fusion protein in transfected LLC-PK 1 cells. LLC-PK 1 cells that were stably transfected with the GFP-pRS1 fusion protein or with GFP alone were grown on glass coverslips, and the distribution of GFP in subconfluent cells was examined by confocal laser-scanning microscopy (CLSM; Fig. 1, a and b ). We found that GFP was evenly distributed throughout the cytosol and the nucleus, whereas GFP-pRS1 was enriched within the nucleus except for the nucleoli. Most cells expressing GFP-pRS1 exhibited high cytosolic concentrations of GFP-pRS1, making it difficult to assess a detailed cytosolic distribution and a putative plasma membrane association of GFP-pRS1 ( Fig. 1 b ). However, in some cells with lower cytosolic concentrations of GFP-pRS1, some perinuclear staining and staining of cell linings were observed ( Fig. 1, c and d ). Figure 1 e shows the Western blot analysis of the subcellular distributions of GFP and GFP-pRS1 in subconfluent LLC-PK 1 cells. GFP was observed exclusively in the cytosolic fraction, detected neither in the PME fraction nor in washed nuclei. In contrast, intact GFP-pRS1 protein was observed in all three fractions (cytosol, PME fraction, and washed nuclei). This indicates that pRS1 directs GFP to the plasma membrane and to the nucleus where it cannot be washed out easily. Note that, in SDS polyacrylamide gels, the 94-kDa GFP-pRS1 polypeptide migrates at 110 kDa. This is consistent with the observation that the 67-kDa polypeptide pRS1 migrates at 100 kDa (see Fig. 2 a and Refs. 16, 32 ).& i$ E+ L, q2 [0 N- i6 a& h
: q+ O0 S7 {% a
Fig. 1. Analysis of green fluorescent protein (GFP) and GFP-pRS1 in transfected LLC-PK 1 cells and in subcellular fractions of LLC-PK 1 cells. a-d : Fluorescence analysis in transfected cells. LLC-PK 1 cell lines that stably expressed GFP ( a ) or GFP-pRS1 fusion protein ( b-d ) were seeded on coverslips and grown for 2 days. The distribution of GFP fluorescence was analyzed by confocal laser-scanning microscopy (CLSM; a and b ) or by epifluorescence ( c ). d : Nomarski image. Bars = 10 µm. Data indicate that pRS1 targets GFP to the nucleus and to the plasma membrane. e : Western blot analysis in subcellular fractions. LLC-PK 1 cells that stably expressed GFP or GFP-pRS1 fusion protein were grown on tissue culture plates to 50% confluence. Cells were homogenized, and plasma membrane-enriched fractions (PME), nuclei, and cytosolic fractions (Cy) were prepared. Samples were separated by SDS-PAGE, blotted to nitrocellulose, and stained with antibody against GFP. Per lane, 10 µg of protein were applied. Data confirm targeting of GFP-pRS1 to the nucleus and to the plasma membrane. ab, Antibody; M r, relative molecular weight.
9 x8 \) t j( L" d+ Z7 u
$ O- O/ \" D3 O. i% ]Fig. 2. Epifluorescence immunolocalization of endogenously expressed pRS1 in LLC-PK 1 cells. a : Western blot in which 1 ng of recombinant pRS1 protein was stained with an affinity-purified antibody that was raised against peptide pRS1-p-ab. Staining could be blocked with the antigenic peptide (ab, blocked). b-i : Staining of subconfluent LLC-PK 1 wildtype cells that were grown on coverslips for 18 h with affinity-purified antibody against total pRS1 protein (pRS1-ab; b-e ) or with affinity-purified antibody against a peptide of pRS1 (pRS1-p-ab; f-i ). Cells were fixed with paraformaldehyde and permeabilized with Triton X-114. They were stained with pRS1-ab or pRS1-p-ab raised in rabbits and a fluorescent (AlexaFluor-555) secondary antibody from goat ( b, d, f, and h ). In d, pRS1-ab has been absorbed with purified recombinant pRS1. In h, pRS1-ab has been blocked by incubation with 0.1 mg/ml antigenic peptide. Counterstaining of the nuclei with 4',6'-diamidino-2-phenylindole (DAPI) in b, d, f, and h is shown in c, e, g, and i, respectively. Arrowheads in b and f indicate immunostaining beneath the plasma membrane converging to presumed cellular attachment areas. Arrows in b and f indicate staining of tubulovesicular structures around the nuclei. Bars = 20 µm. Data indicate location of endogenous pRS1 in nuclei, in a compartment around the nuclei, and at the plasma membrane.
) ]( T+ q, T' ?! ^( w3 c
M% m3 q, M' A( N. }% D" W1 kSubcellular distribution of endogenous pRS1 in LLC-PK 1 cells. Subcellular localization of endogenous pRS1 protein in subconfluent LLC-PK 1 wildtype cells was studied by immunofluorescence microscopy using a previously described antibody against total pRS1 protein (pRS1-ab) ( 16, 32 ) and a new antibody raised against a peptide of the central portion of pRS1 (pRS1-p-ab). As shown by Western blotting, both affinity-purified antibodies bound to recombinant pRS1 protein, and the reaction was blocked by pRS1 protein and by antigenic peptide, respectively (for pRS1-ab, see Ref. 32; for pRS1-p-ab, see Fig. 2 a ). Previous Western blots showed that pRS1-ab bound specifically to pRS1 in plasma membranes of porcine renal proximal tubules and of LLC-PK 1 cells ( 16, 32 ). Subconfluent LLC-PK 1 cells were stained after culture on coverslips ( Fig. 2, b-i ). Both antibodies against pRS1 stained intracellular structures beneath the plasma membrane ( Fig. 2, b, c, f, and g ). The staining intensity was increased in regions near focal attachment points (arrowheads in Fig. 2, b and f ), and prominent staining was observed in a compartment around the nuclei (arrows in Fig. 2, b and f ). In subconfluent cells, most nuclei, as identified by counterstaining with DAPI, were stained ( Fig. 2, b, c, f, and g ). The specificity of the immunoreactions was verified by preabsorption of pRS1-ab with pRS1 protein ( Fig. 2, d and e ) and of pRS1-p-ab with antigenic peptide ( Fig. 2, h and i ). Employing CLSM, further details on the intracellular localization of RS1 could be distinguished ( Fig. 3 ). Immunoreactivity of RS1 below the plasma membrane and within the cytosol could be assigned to small vesicles (see arrowheads in Fig. 3, a and b ), and staining around the nuclei could be attributed to membrane stacks and vesicles (arrows in Fig. 3 c ).& n8 x4 \/ ?% ^3 d6 U* ^6 E
0 Q/ \5 @1 L) j3 o5 L- lFig. 3. Immunolocalization of pRS1, pSGLT1, and microtubules in LLC-PK 1 cells by CLSM. a-c : Localization of pRS1. Nontransfected LLC-PK 1 cells were grown on coverslips to 50% confluence, fixed, permeabilized, and incubated with DAPI and/or antibodies as in Fig. 2. Incubation was performed with affinity-purified pRS1-ab followed by incubation with anti-rabbit IgG F(ab') 2 linked to the red fluorescent dye AlexaFluor-555 (red). Bars: 10 µm ( a ) and 2 µm ( b and c ). Analysis by CLSM revealed further details about the intracellular distribution of pRS1. In addition to the plasma membrane and the nucleus, immunoreactivity of pRS1 was observed at vesicles below the plasma membrane and within the cytosol (arrowheads in a and b ) and at tubulovesicular structures and vesicles around the nucleus (arrows in c ). d-i : Parallel staining of pSGLT1 or pRS1 with -tubulin. Nontransfected subconfluent LLC-PK 1 cells grown on coverslips were fixed, permeabilized, and stained with mouse monoclonal antibody against -tubulin ( d-i ) and rabbit anti-SGLT1 antibody QIS30 ( d-f ) or rabbit pRS1-ab ( g-i ). Anti-rabbit IgG F(ab') 2 linked to AlexaFluor-555 (red) and anti-mouse IgG linked to Cy2 (green) were used as secondary antibodies. In f and i, nuclei were stained with DAPI (blue). Bars = 10 µm. Confocal images show that SGLT1 is located in tubulovesicular structures that are distributed throughout the cells and are often colocated with microtubules (arrowheads in e and f ). pRS1 is located at tubulovesicular structures around the nuclei and at small vesicles in the cytosol (arrows in g ). For some vesicles and tubulovesicular structures, colocalization with microtubules was observed (arrowheads in h and i ).' ^8 ^3 w) O, l% T1 |% J& Z
: b' b+ A6 r9 @4 K) ]5 GComparison of subcellular distributions of porcine SGLT1 and porcine RS1 in relation to microtubules. To get an idea where posttranscriptional regulation of SGLT1 by RS1 takes place, we tried to identify common intracellular localizations of SGLT1 and RS1. Because previous work demonstrated SGLT1 in endosomal compartments of Caco-2 cells that were distributed along microtubuli ( 13, 14 ), we performed parallel staining of -tubulin with either SGLT1 or RS1 in subconfluent LLC-PK 1 cells. For staining of SGLT1, we used the previously described antibody QIS30 ( 14 ). QIS30 was raised against an intracellular domain of SGLT1 from rabbit and cross-reacts with human SGLT1 ( 14 ) and with porcine SGLT1 (pSGLT1) (data not shown). To allow intracellular staining, cells were permeabilized with Triton X-114. In subconfluent LLC-PK 1 cells, immunoreactivity of SGLT1 ( Fig. 3, d and f; red) was observed in tubulovesicular compartments that were distributed throughout the cytosol, showing some concentration within the perinuclear field. Parallel staining of SGLT1 and -tubulin suggested that the tubulovesicular structures containing SGLT1 are lined up along microtubules ( Fig. 3, e and f; see arrowheads), similar to the distribution observed in Caco-2 cells ( 14 ). These tubulovesicular structures containing SGLT1 exhibited short diameters of 1.0 ± 0.2 µm (mean value ± SD, n = 103). Because of permeabilization with Triton X-114, the plasma membrane localization of SGLT1 was not visible, as observed earlier in Caco-2 cells ( 14 ). Parallel staining for pRS1 and -tubulin yielded different results ( Fig. 3, g-i ). In addition to the above mentioned perinuclear compartment, vesicles in the cell periphery were stained (see arrows in Fig. 3 g ), the diameters of which were smaller compared with the diameters of vesicles with SGLT1 or the small diameters of tubulovesicular compartments with SGLT1 (0.35 ± 0.13 µm; mean value ± SD, n = 101, P
' e9 m @+ y" M& M5 M- I T; D3 e
) b; S+ y: _. O9 c" zTo further investigate the role of microtubules for the cytosolic location of SGLT1 and RS1 immunoreactivity, we determined the distribution of both proteins after treatment of cells with colchicine, a microtubule-depolymerizing toxin. Subconfluent LLC-PK 1 cells were incubated with 10 µM colchicine for 2 h. Staining of the cells with antibody against -tubulin showed that nearly all microtubules were depolymerized after this period of time (data not shown). By colchicine treatment, the orientation of tubulovesicular structures containing SGLT1 in the cytosol was abolished ( Fig. 4, compare a and b ). At variance, colchicine treatment led to a loss of the tightly perinuclear location of the tubulovesicles containing RS1. They were redistributed throughout the entire cytosol ( Fig. 4, compare c and d ). RS1 staining of the plasma membrane and of small vesicles below the plasma membrane was not changed significantly by colchicine (data not shown). The data support the interpretation that the cytosolic tubulovesicular compartments containing SGLT1 are associated with microtubules. They suggest a location of RS1 at the Golgi complex that requires intact microtubules for organization.
2 O; R- g# Y, q7 L) `+ p+ ^
6 e! }' A, F, z. VFig. 4. Effects of colchicine on distribution of SGLT1 and RS1 in LLC-PK 1 cells. Nontransfected LLC-PK 1 cells were grown on coverslips up to 50% confluence ( a-d ). Some of them were incubated for 2 h with 10 µM colchicine (COL; b and d ). Cells were stained for SGLT1 with antibody QIS30 ( a and b ) or for RS1 with antibody pRS1-ab ( c and d ). AlexaFluor-555-coupled anti-rabbit IgG F(ab') 2 was used as secondary antibody, and nuclei were counterstained with DAPI. Staining was visualized by epifluorescence microscopy. Bars = 10 µm. In this picture, staining of RS1 at the plasma membrane has not been resolved. Colchicine destroyed the orientation of the SGLT1-containing tubulovesicular compartment within cytosol as well as the organization of perinuclear tubulovesicular stacks with RS1 immunoreactivity.
( ~- N, c' D; X" d: y
( e& b1 P* {- w2 l" g @0 i2 ~Assignment of RS1 and SGLT1 to the TGN. To demonstrate the location of RS1 at the Golgi complex and, more precisely, at its substructure, the TGN, and to determine whether SGLT1 can be detected at the Golgi complex, we performed parallel staining experiments for both proteins with a TGN marker protein. Because no antibody against a porcine TGN marker protein was available, these experiments were conducted with HEK-293 cells. Like LLC-PK 1 cells, HEK-293 cells express both SGLT1 and RS1 endogenously. This was shown in Western blots ( Fig. 5 ) and by immunostaining ( Fig. 6 ). In HEK-293 cells, sometimes degradation products of RS1 were observed in addition to intact hRS1 protein (data not shown). For identification of the TGN, we used an antibody against TGN46, the human ortholog of TGN38 from rat ( 1, 18 ). The intracellular distributions of SGLT1 and RS1 in HEK-293 cells were similar to those in LLC-PK 1 and Caco-2 cells (compare Fig. 6 with Fig. 3 and with Ref. 14, Fig. 4 ) with the exception that the association of RS1 with the plasma membrane was difficult to resolve in HEK-293 cells. Parallel staining of SGLT1 and TGN46 showed that only a small fraction of the TGN46-positive compartment contained SGLT1 ( Fig. 6, a-c, arrowheads). In contrast, staining of RS1 and TGN46 in the same cells showed that RS1 is located at the entire TGN46-positive compartment ( Fig. 6, d-i, arrowheads). The data indicate that RS1 is located at the entire TGN and accordingly must be colocated with SGLT1 at parts of the TGN. This colocation of RS1 and SGLT1 is consistent with the hypothesis that RS1 inhibits the membrane expression of SGLT1 posttranscriptionally at the TGN.) J2 t( q, b: e8 Y+ O
9 e8 k0 S/ ?4 J( h
Fig. 5. Western blots showing the expression of RS1 and SGLT1 in human embryonic kidney (HEK)-293 cells. Nontransfected HEK-293 cells ( lanes 1 and 3 ) and HEK-293 cells transiently transfected with pRS1 ( lane 2 ) were grown to 80% confluence. Cells were harvested, and PME membrane fractions were isolated. PME fractions (20 µg of proteins per lane) were separated by SDS-PAGE, transferred to polyvinyldifluoride membrane, and incubated with affinity-purified antibody against pRS1 (pRS1-ab, diluted 1:2,500; lanes 1 and 2 ) or with affinity-purified antibody against SGLT1 (QIS30, diluted 1:4,000; lane 3 ). Blots show the presence of endogenous and expressed RS1 protein and endogenous SGLT1 protein in HEK-293 cells.2 u8 F% l; U9 K4 O
7 }/ L' e Y8 j4 c) @/ }# h
Fig. 6. Demonstration that RS1 is associated with the entire trans -Golgi network (TGN), whereas SGLT1 is only present in some areas of the TGN. HEK-293 cells were grown on coverslips to 50% confluence, fixed, and permeabilized. Cells were stained either with an antibody (QIS30) directed against SGLT1 that was raised in rabbit plus an antibody directed against human TGN46 that was raised in sheep ( a-c ) or with RS1-ab raised in rabbit and antibody against human TGN46 ( d-i ). Anti-rabbit IgG F(ab') 2 linked to AlexaFluor-555 (red) and anti-sheep IgG linked to Cy2 (green) were used as secondary antibodies. Nuclei were stained with DAPI (blue). Fluorescence staining was analyzed by CLSM. Bars: 10 µm ( a-f ) and 2 µm ( g-i ). RS1 exhibits a virtually complete colocation with TGN46 (arrowheads in d-i ). For SGLT1 and TGN40, colocation was observed only in restricted areas (arrowheads in a-c ).: U4 `1 A- t* z" j0 ~
7 h# L4 h0 H; a+ ?( q9 L5 MParallel staining of RS1 with clathrin or dynamin II or of SGLT1 with clathrin or dynamin II. Previously, we showed that expression of Na - D -glucose cotransport after injection of SGLT1 cRNA into oocytes of X. laevis was inhibited by coinjection of RS1 cRNA, and that this posttranscriptional inhibition was abolished when a dominant-negative mutant of dynamin was expressed in the same oocytes ( 17, 32, 35 ). To try to distinguish whether RS1 increases dynamin-dependent endocytosis of SGLT1-containing vesicles at the plasma membrane or inhibits dynamin-dependent vesicle release from the TGN, we investigated whether RS1 and SGLT1 are colocated with dynamin and clathrin. As shown in Fig. 7, we performed parallel immunostaining of RS1 and SGLT1 with clathrin and dynamin II. As expected, a broad intracellular distribution was observed for clathrin and dynamin with the nuclei containing neither clathrin nor dynamin. At the plasma membrane, no significant colocation between RS1 and clathrin or RS1 and dynamin was observed (see arrows in Fig. 7, a, c, d, and f, for RS1 staining; see open arrowheads in Fig. 7, b and c, for staining of clathrin; and see open arrowheads in Fig. 7, e and f, for staining of dynamin). In the perinuclear compartments, partial overlap was observed for staining of RS1 and clathrin (solid arrowheads in Fig. 7 c ) and for staining of RS1 and dynamin (solid arrowheads in Fig. 7 f ). Parallel immunostaining of SGLT1 and clathrin revealed virtually no colocalization of these two proteins ( Fig. 7, g-i ). At variance, colocation of SGLT1 and dynamin in the perinuclear area was observed in small punctuate areas (solid arrowheads in Fig. 7, j-l ). The data are consistent with the hypothesis that RS1 inhibits the dynamin-dependent release of SGLT1-containing vesicles from the TGN.2 Q5 b! I. A4 B9 R' ^. B0 i
# \/ s, N0 ]4 h+ y* Y) M; k! FFig. 7. Colocation of RS1 with clathrin and dynamin at the TGN but not at the plasma membrane. LLC-PK 1 cells were grown on coverslips to 50% confluence, fixed, and permeabilized. Cells were incubated either with pRS1-ab plus monoclonal antibody against clathrin (CLA) from bovine brain raised in mice ( a-c ), with RS1-ab and a peptide antibody against human dynamin II (DYN) raised in goat ( d-f ), with antibody QIS30 against SGLT1 plus antibody against clathrin ( g-i ), or with antibody against SGLT1 plus antibody against dynamin II ( j-l ). Secondary antibodies were goat-anti-rabbit IgG linked to AlexaFluor-555 (red, RS1 or SGLT1), goat-anti-mouse IgG linked to Cy2 (green, CLA), and donkey-anti-goat IgG linked to AlexaFluor-488 (green, DYN). Fluorescence staining was analyzed by CLSM. Bars: 10 µm ( a-f ) and 5 µm ( g-l ). RS1 is colocated with clathrin and dynamin at distinct regions of the TGN (yellow staining and solid arrowheads in c and f ). At the plasma membrane, no significant colocation of RS1 with clathrin of dynamin was observed (open arrowheads indicating green plasma membrane areas in b, c, e, and f, and arrows indicating red plasma membrane areas in a, c, d, and f ). SGLT1 shows no colocation with clathrin but colocates in distinct areas with dynamin (arrowheads in j-l ).
7 A, u C4 K+ \- @" T4 E# y" a8 l) Z5 q; A
BFA induces disappearance of RS1 from the TGN. BFA is a fungal metabolite that has been used extensively to decipher vesicular transport processes in eukaryotic cells ( 15 ). The most striking effects of BFA are the release of various coat proteins from the Golgi apparatus and morphological changes of intracellular tubulovesicular compartments that reflect changes in membrane trafficking pathways. Targets of BFA are guanosine nucleotide exchange factors (GEFs) that activate ADP-ribosylation factors (ARFs), which regulate the assembly of vesicle coat complexes on the TGN ( 3, 5, 6, 9, 10, 26 ). To determine whether RS1 is a GEF/ARF-dependent coat protein at the TGN, we incubated subconfluent LLC-PK 1 cells for various time periods with 2 µg/ml BFA and performed immunostaining for SGLT1 and RS1 ( Fig. 8 ). After 1- or 5-min incubation of subconfluent LLC-PK 1 cells with BFA, distinct morphological changes of the tubulovesicular compartments with SGLT1 immunoreactivity were observed. The relatively close packing of tubulovesicular compartments with SGLT1 observed in many cells became more dissociated, and increasing numbers of single tubules with extensive ramification became apparent ( Fig. 8, a-c ). SGLT1 remained within the intracellular membranes. In contrast, the immunoreactivity of RS1 at the perinuclear compartment disappeared within several minutes after incubation of the LLC-PK 1 cells with BFA. The data show that RS1 protein is released from the TGN by BFA and suggest that RS1 is a GEF/ARF-dependent protein that coats the TGN.- `( W5 J) E+ S b: w) y
5 d7 A& P* ^: d" Y/ F+ eFig. 8. Brefeldin A (BFA) induces disappearance of RS1 from the perinuclear compartment in LLC-PK 1 cells. Subconfluent LLC-PK 1 cells grown on coverslips were incubated with 0.1% methanol (controls, a and d ) or with 2 µg/ml BFA dissolved in 0.1% methanol ( b, c, e, and f ). Incubation with BFA was performed for 1 min (BFA 1) or 5 min (BFA 5). Cell metabolism was stopped by transfer of the cells on ice and superfusion with cold washing buffer. After paraformaldehyde fixation and permeabilization, the cells were probed with antibody against SGLT1 ( a-c ) or with pRS1-ab ( d-f ). Goat-anti-rabbit IgG F(ab') 2 linked to AlexaFluor-555 was used as secondary antibody. Bars = 5 µm. In these epifluorescence images, a complete disappearance of RS1 from the perinuclear compartment was observed, which was almost complete after 5 min.$ Q' ]5 b1 ~* M. `' x/ Z; [2 u) O' r
, i- ]: C5 `5 y u" { O7 U9 C) j+ i
Changes of intracellular distribution of RS1 and SGLT1 during confluence. During confluence of LLC-PK 1 cells, the intracellular concentration of RS1 is downregulated, whereas SGLT1 is transcriptionally upregulated ( 16 ). It has been demonstrated that the transcriptional inhibition of SGLT1 by RS1 is responsible for the downregulation of SGLT1 in subconfluent cells and that the upregulation of SGLT1 after confluence is due to a relief of transcriptional inhibition by RS1 ( 16 ). Wondering whether a confluence-dependent redistribution of RS1 may be involved in this regulation, we compared the cellular distribution of RS1 before and after confluence. For these experiments, LLC-PK 1 cells were grown on polyester membranes. In subconfluent cells grown on filters, the same staining for RS1 and SGLT1 was obtained as in subconfluent cells grown on coverslips (for RS1, compare Fig. 2, b and c, with Fig. 9, a and b ). With pRS1-ab, most nuclei, the plasma membrane, and the perinuclear compartment were stained. At variance, when the cells were grown for 8 more days after reaching confluence, no specific immunostaining for pRS1 could be detected at plasma membranes and within nuclei ( Fig. 9, c and d ). In these cells, RS1 immunoreactivity in the perinuclear compartment could be readily resolved by CLSM ( Fig. 9 e ). The immunostaining for RS1 before and after confluence suggested a reduction of total RS1 protein in LLC-PK 1 cells during confluence. This is consistent with previous data showing that a crude membrane fraction isolated from LLC-PK 1 cells contained much less RS1 protein when it was isolated from confluent vs. subconfluent cells (see Fig. 2 C in Ref. 16 ). In the same study, an inverse relationship was obtained for SGLT1. In crude membranes isolated from confluent LLC-PK 1 cells, much more SGLT1 protein was observed compared with subconfluent cells (see Fig. 2 A in Ref. 16 ). Consistent with this observation, we noted a much stronger immunoreactivity for SGLT1 in confluent LLC-PK 1 compared with subconfluent cells.# B F+ G0 ?0 T+ p2 f9 F
1 l; R/ I8 s% g: h$ W wFig. 9. Effects of cell confluence and inhibition of the proteasome on the distribution of RS1 and SGLT1 in LLC-PK 1 cells. LLC-PK 1 cells were seeded on polyester membranes and grown for 18 h after seeding ( a and b ) or for 8 days after reaching confluence ( c-h ). In g and h, cells were incubated for 16 h with 10 µM MG132. Cells were permeabilized with detergent and stained with affinity-purified rabbit pRS1-ab ( a, c, e, and g ), affinity-purified rabbit antibody against SGLT1 ( f and h ), and a fluorescent secondary antibody from goat. Counterstaining of the nuclei with DAPI in a and c is shown in b and d, respectively. Bars: 20 µm ( a and c ) and 5 µm ( e and g ). In i, LLC-PK 1 cells were seeded on polyester membranes and grown for different time intervals. Cells were permeabilized, immunostained for RS1, and counterstained with DAPI. Per time interval and experiment, 1,000-8,000 nuclei were identified by DAPI fluorescence and evaluated for immunostaining of RS1, and the fractions of the immunostained nuclei were calculated. Mean values ± SD of 3 independent experiments each are indicated.
( g$ p8 q+ Y/ Q1 \2 [8 N- _5 u; j7 K3 e7 B
To quantify the nuclear localization of endogenous pRS1 in relation to cell confluence, we seeded a fixed number of LLC-PK 1 cells on polyester membranes, grew them for up to 13 days whereby confluence occurred at the fifth day, stained them with pRS1-ab, and counted the percentage of immunoreactive nuclei. In subconfluent cells, 90 ± 6% of the nuclei were stained ( Fig. 9 i ). The number of stained nuclei was decreased before and during confluence. Four to eight days after reaching confluence, only between 5 and 20% stained for pRS1. To decide whether the absence of RS1 immunoreactivity in the nuclei observed after confluence results from the downregulation of the amount of pRS1 protein in the cytosol or whether the nuclear migration of RS1 is confluence dependent, we also measured confluence-dependent nuclear localization of pRS1 after overexpression of GFP-pRS1 fusion protein. After transient transfection of LLC-PK 1 cells, we seeded them onto polyester membranes and grew them either to 50% confluence or until 4 days after confluence. In three separate experiments, we determined the fraction of transfected cells that contained GFP-pRS1 within their nuclei (90-150 cells per experiment). Before confluence, 84 ± 3% of the transfected cells, and after confluence, 32 ± 10%, showed significant nuclear staining ( P 2 R7 I2 T8 i. W3 x
" j, Z& A) \$ r8 ^4 aBecause pRS1 does not contain a known classical or nonconventional nuclear localization sequence ( 8, 29, 36, 37, 38 ) and our data indicated confluence-dependent regulation of nuclear migration of RS1, we wondered whether nuclear migration of RS1 is dependent on the amount or activity of a cytosolic interacting protein or inhibitor. In a survey experiment, we increased the concentrations of proteins that undergo proteasomal degradation in confluent LLC-PK 1 cells by incubating them for 16 h with 10 µM proteosomal inhibitor MG132 and investigated RS1 immunoreactivity in the nuclei. After treatment with MG132, overall immunoreactivity of RS1 was increased, suggesting that RS1 itself is degraded by the proteasome ( Fig. 9, compare e and g ). A similar effect of MG132 was observed in Western blots of cell lysates from subconfluent HEK-293 cells that were transiently transfected with pRS1. After incubation of the transfected cells for 16 h with 10 µM MG132, the amount of intact pRS1 protein in the cytosol was increased to 232 ± 25% of the control without MG132 as revealed by Western blots (mean ± SD, n = 3, P " p3 B5 H$ ]2 E, U8 W* B
7 P9 Q% B: l- e7 d& q
In the experiments depicted in Fig. 9, f and h, we investigated the effect of MG132 on expression and distribution of SGLT1 protein in confluent LLC-PK 1 cells. As expected, the amount of SGLT1 in the cytosol was decreased by MG132. This is probably due to the increased amount of pRS1. In addition, the intracellular distribution of SGLT1 was changed. Whereas in untreated confluent LLC-PK 1 cells SGLT1 is distributed in tubulovesicular compartments throughout the cytosol, after treatment with MG132, SGLT1 protein was restricted to a perinuclear compartment. A similar effect of MG132 on the cytosolic distribution of SGLT1 was observed in subconfluent LLC-PK 1 cells (data not shown). These observations are consistent with the hypothesis that RS1 at the TGN has an inhibitory effect on the sorting of SGLT1 in the TGN or on release of SGLT1-containing vesicles from the TGN. Since, in subconfluent LLC-PK 1 cells that express high amounts of RS1 in the absence of MG132, SGLT1 is distributed in tubulovesicular compartments throughout the cytosol ( Figs. 3 d and 6 a ), high amounts of intact RS1 alone are not sufficient to explain the effect of MG132 on the intracellular distribution of SGLT1. Most likely, additional proteins are involved in the sorting of SGLT1 and/or release of SGLT1-containing vesicles and are upregulated in the presence of MG132. One candidate is the recently identified ischemia/reperfusion-inducible protein IRIP ( 11 ).7 u: h( `# R% _9 z* m7 X
6 Y0 D+ M7 |" ?3 z1 Y- fDISCUSSION
% c0 _7 u' I/ ]" Y: s
* r, R: \7 X( S6 l3 WTo better understand the intracellular functions of RS1, we determined its intracellular locations in confluent and subconfluent epithelial cells compared with one of its target proteins, the Na - D -glucose cotransporter SGLT1. RS1 has been shown previously to be of high physiological and biomedical interest because it is involved in the regulation of glucose absorption in the small intestine ( 21 ) and of renal transporters after ischemia and reperfusion ( 11 ). RS1 also plays important roles in the widely used model system of LLC-PK 1 cells for the downregulation of SGLT1 in the subconfluent state. RS1 is associated with the 28-kDa protein IRIP, which is upregulated in kidney after ischemia and reperfusion ( 11 ). IRIP inhibits the expression of a variety of plasma membrane transporters such as the organic cation transporter OCT2, which has been shown to be inhibited by RS1 as well ( 35 ). Because inhibition of OCT2 by RS1 and IRIP was not additive, and inhibition of OCT2 by RS1 was prevented by coexpression of a dominant-negative mutant of IRIP, RS1 and IRIP are supposed to be parts of a common regulatory pathway controlling transporter activities. K2 ?; ~: i% u$ @" {& q
: T7 M2 I$ n& }3 n+ P7 ]: C |Previous experiments indicated that RS1 inhibits the expression of SGLT1 both transcriptionally and posttranscriptionally and that the posttranscriptional inhibition of SGLT1 by RS1 is dependent on the function of dynamin ( 16, 21, 35 ). In the present paper, we show that the RS1 protein is located at the intracellular side of the plasma membrane, at vesicles below the plasma membrane, at the TGN, and within the nucleus. In comparing the distribution of RS1 protein and SGLT1 protein, considerable differences were observed. However, both proteins are colocated at the TGN together with dynamin, and are most probably also colocated at the plasma membrane. Colocation of RS1 and SGLT1 at the plasma membrane could not be demonstrated by immunofluorescence microscopy because the plasma membrane had to be permeabilized to detect the intracellular RS1 protein, and this destroyed SGLT1 immunoreactivity within the plasma membrane. However, because RS1 was found at the entire plasma membrane, colocation of RS1 and SGLT1 at the plasma membrane is probable. Intracellularly, most SGLT1 was associated with tubulovesicular compartments previously identified as endosomes in Caco-2 cells ( 14 ) that were lined up along microtubules. RS1 protein was found at smaller vesicles but mostly at tubulovesicular structures of the TGN. When the microtubules were destroyed with colchicine, the tubulovesicular compartments with SGLT1 obtained a more random distribution, and the tubulovesicles with RS1 distributed throughout the cytosol in accordance with the requirement of microtubules for the integrity of the Golgi complex ( 19, 24, 25, 31 ). The observed colocation of RS1, SGLT1, and dynamin at the TGN is consistent with the hypothesis that RS1 blocks the release of SGLT1 from the TGN. RS1 either inhibits the dynamin-dependent release of SGLT1-containing vesicles from the TGN or the sorting of SGLT1 into these vesicles. It is noteworthy that SGLT1 did not colocate with clathrin perinuclearly. For this reason, it is probable that SGLT1 distributes into vesicles at the TGN that are not clathrin coated. The hypothesis that RS1 inhibits SGLT1 at the TGN is supported by functional experiments described in the accompanying paper by Veyhl et al. ( 33a ).3 d; B8 m) C* ?: C2 M
( q! O! g$ o( ~
The BFA-induced dissociation of RS1 from the TGN suggests that RS1 belongs to proteins that form coat complexes or regulate their formation at the TGN and may be involved in TGN sorting. BFA inhibits GEFs that activate ARFs, which regulate the assembly of coat complexes at the TGN ( 3, 5, 6, 9, 10, 26 ). According to the current model, activated ARFs recruit coat proteins such as Golgi-localized -ear-containing ARF-binding proteins (GGAs). These coat proteins serve as adaptors for "cargo" proteins that are directed to endosomes or lysosomes. It has been shown that domains of the GGAs bind to an acidic-cluster-dileucine motif in mannose-6-phosphate receptors and thereby target mannose-6-phosphate receptors to endosomes ( 7, 27 ). Interestingly, GGAs also bind to ubiquitinated proteins at the TGN, and RS1 contains a ubiquitin binding-associated (UBA) domain that binds tetraubiquitin (Müller T and Koepsell H, unpublished data). The UBA domain of RS1 contains a DLALL motif that is conserved in RS1 from different species and meets the minimal requirements of the acidic-cluster-dileucine motif ( 20, 30 ). A recent report suggested that SGLT1 is ubiquitinated by the ubiquitin ligase Nedd4-2 and that ubiquitination of SGLT1 plays an important role in the turnover of SGLT1 ( 4 ). It is therefore tempting to speculate that RS1 blocks the sorting of ubiquitinated transporters at the TGN or the release of vesicles from the TGN by binding to ubiquitinated cargo proteins in the TGN and/or adapter proteins (e.g., GGAs) that bind to the acidic-cluster-dileucine motif.' `- ^ |/ n" \/ g/ h. D% i! E
+ z8 R7 I# w1 ?; R
An interesting observation of the present paper was that the nuclear localization of RS1 in subconfluent LLC-PK 1 cells was drastically decreased after cell confluence. Previously, we observed that the cytosolic concentration of RS1 was high in subconfluent cells and low in confluent cells, whereas transcription of SGLT1 and concentration of mRNA were low before confluence and high after confluence ( 16 ). This inverse relationship between RS1 and SGLT1 gave rise to the hypothesis that RS1 suppresses transcription of SGLT1 in subconfluent cells and that the upregulation of SGLT1 after confluence is caused by a relief of this inhibition. The nuclear location of RS1, as described above, suggests that RS1 can directly interact with the transcriptional complex of SGLT1. The confluence dependence of the nuclear localization suggests that the confluence-dependent inhibition of SGLT1 transcription is partially regulated via nuclear migration. Recently, we identified a new, nonconventional, 21-amino acid-long nuclear targeting sequence in RS1 (NLS-R) that is framed by a consensus sequence for casein kinase II at the NH 2 terminus and by consensus sequences for PKC phosphorylation at the COOH terminus. Together with these phosphorylation sites, the nuclear targeting sequence of RS1 mediates confluence-dependent nuclear targeting (Leyerer M, Gorboulev V, Filatova A, Kroiss M, Müller TD, and Koepsell H, unpublished data). We suggest that proteins not yet identified participate in nuclear migration of RS1 and that kinases regulate the confluence dependence of nuclear migration.
$ }. [: v! O" y0 m R7 l3 }" u9 R+ W* f- g9 N
RS1 and the recently identified protein IRIP belong to a novel pathway that regulates transporters of several families. Because RS1 and IRIP are expressed in various tissues, the regulatory pathway is supposed to be present in many cell types. The observed confluence dependence of RS1 concentration and nuclear migration of RS1 and the upregulation of IRIP after ischemia and reperfusion in kidneys suggest that this pathway is involved in differentiation/dedifferentiation in response to various physiological stimuli. The detection of the different intracellular locations of RS1 and their dynamic changes in relation to cell confluence provided important new insights for the understanding of this pathway, in which transcriptional and posttranscriptional regulations appear to be coordinated. Additional functional experiments, the identification of all components of this pathway, and detailed characterizations of mice with deleted RS1 and IRIP will help in better understanding the functional mechanism and physiological relevance of this pathway.) z2 K- j0 |. `" Z2 U
/ c# T+ ~( U, @5 f& z9 r* D
GRANTS
1 x+ G# C0 V4 {, \! q6 c# |3 }, v. ]! W% B7 l! R' u, [
This work was supported by the Deutsche Forschungsgemeinschaft Grant SFB 487/C1.
3 F* q$ V8 y2 u9 u, J
! q; z8 ?6 j. z$ XACKNOWLEDGMENTS
* w/ G5 X4 G; e; ^3 I# G6 h4 Y
; ?6 G) m g% ^( s" v: s& cLaser-scanning microscopy was performed at the Institut für Molekulare Infektionsbiologie of the University of Würzburg. Figures were prepared by M. Christof." P0 a, d( A5 z
【参考文献】
" L$ X; t) }2 y$ T- w, |7 W& [9 u Banting G and Ponnambalam S. TGN38 and its orthologues: roles in post-TGN vesicle formation and maintenance of TGN morphology. Biochim Biophys Acta 1355: 209-217, 1997.: g$ {; h2 z% {) \/ E% M! |
: M0 k% I7 ^8 O) b6 x
, H, A" Z0 t6 Z' @% R% ~+ k3 j, Y/ |) \8 ?7 Z0 g( s
Bradford MM. A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72: 248-254, 1976.
1 `; C D1 k/ X R( V% A) v( m8 r$ ]4 O1 }; A! a: k z
/ z0 j* M5 \3 u- N3 v: ?& w, A4 T( q. d" Z v* W! _4 W- P
Chardin P and McCormick F. Brefeldin A: the advantage of being uncompetitive. Cell 97: 153-155, 1999.
7 {/ B1 x3 o0 l" F& ~$ O1 x5 d; d; R
! E; s. [0 I9 i+ r& q7 _' {( G0 Q' R- k! o& M
Dieter M, Palmada M, Rajamanickam J, Aydin A, Busjahn A, Boehmer C, Luft FC, and Lang F. Regulation of glucose transporter SGLT1 by ubiquitin ligase Nedd4-2 and kinases SGK1, SGK3, and PKB. Obes Res 12: 862-870, 2004.
2 m! ~4 T& ~3 w* v! _! T; z% M0 `6 ?8 K3 U# W
) Z, c. H$ n! q3 l
, ~; ~. }4 ] w
Dinter A and Berger EG. Golgi-disturbing agents. Histochem Cell Biol 109: 571-590, 1998.* e* g6 f" K( c2 ]' \. ?& n1 B
$ `' \8 C5 M3 M# Z% p' W
" m2 _6 H% C" p( U! r, \0 {7 y8 j" Y; f
Donaldson JG, Finazzi D, and Klausner RD. Brefeldin A inhibits Golgi membrane-catalysed exchange of guanine nucleotide onto ARF protein. Nature 360: 350-352, 1992.
3 Q$ B( s* l+ D- S$ b# Y: t
% K. T/ k4 b9 r7 B- N0 ]' O
5 i, ~9 ^7 ?) ?4 ~0 U5 ?3 f" Z' P& S4 f2 N/ t
Ghosh P and Kornfeld S. The GGA proteins: key players in protein sorting at the trans -Golgi network. Eur J Cell Biol 83: 257-262, 2004.8 D: l& g* R! f3 N2 C
2 Q' I* m6 r5 O+ X
) f6 L) C+ A! L8 R
: D8 R/ K S! f0 B/ W sGrozinger CM and Schreiber SL. Regulation of histone deacetylase 4 and 5 and transcriptional activity by 14-3-3-dependent cellular localization. Proc Natl Acad Sci USA 97: 7835-7840, 2000.* ^' h# M' k- }$ A- m7 ~; s
4 q4 N$ T2 Z1 c& I# L8 Q* Z+ h
. F% d; {; N: L, |4 i! t! K
: \& c) ], {) N% F2 PHelms JB and Rothman JE. Inhibition by brefeldin A of a Golgi membrane enzyme that catalyses exchange of guanine nucleotide bound to ARF. Nature 360: 352-354, 1992.7 M9 C. u) P! j
7 `6 t7 }2 k/ |
. i0 s% I/ o8 I: Y! @6 d
g% w: }; w3 ^7 i Y4 pJackson CL and Casanova JE. Turning on ARF: the Sec7 family of guanine-nucleotide-exchange factors. Trends Cell Biol 10: 60-67, 2000.
; H& w9 D2 `% g. w% v6 F
, K, p1 `4 A1 ?9 i! s% E: }. N' |& Q9 h" `: X% C! P
- S- l# d; x! q9 k8 n
Jiang W, Prokopenko O, Wong L, Inouye M, and Mirochnitchenko O. IRIP, a new ischemia/reperfusion-inducible protein that participates in the regulation of transporter activity. Mol Cell Biol 25: 6496-6508, 2005." \( v; S. A# ~
* [) n, ]: M( d9 J
7 I x, ^0 ~, G. V! {4 L
/ @# Y& N, |5 VJones SM, Howell KE, Henley JR, Cao H, and McNiven MA. Role of dynamin in the formation of transport vesicles from the trans -Golgi network. Science 279: 573-577, 1998.$ G! Z4 U+ c& f+ N
- c5 F+ q* b& L& F
2 o8 @- B1 m8 Z
} x! o$ w# U5 d1 f* X9 A" A1 NKhoursandi S, Scharlau D, Herter P, Kuhnen C, Martin D, Kinne RKH, and Kipp H. Different modes of sodium- D -glucose cotransporter-mediated D -glucose uptake regulation in Caco-2 cells. Am J Physiol Cell Physiol 287: C1041-C1047, 2004." `, W m& M9 V
8 _7 Z" m% T$ c! ? z) M3 H' E
( ]+ x( l" F0 [! `: @8 ^
% ]) M$ K: o3 k6 H; ~" E l/ i0 m$ o8 ]Kipp H, Khoursandi S, Scharlau D, and Kinne RKH. More than apical: distribution of SGLT1 in Caco-2 cells. Am J Physiol Cell Physiol 285: C737-C749, 2003.
, A; f/ H0 k* j1 ]* ?" |( [9 ?
, j5 a/ R3 o; U+ ]- _4 G9 h) k2 {4 P) h; k- s" O2 z: G" N
' I( U& n' w% W5 J5 K6 `8 A8 ^
Klausner RD, Donaldson JG, and Lippincott-Schwartz J. Brefeldin A: insights into the control of membrane traffic and organelle structure. J Cell Biol 116: 1071-1080, 1992.
$ L. n" Z5 r7 s" m5 W/ J9 G
, e4 ~( F3 X) t; H, A+ i) U0 U1 q
5 a# \, e* c; p2 a% M% A" c$ N- v2 \: W4 m
Korn T, Kühlkamp T, Track C, Schatz I, Baumgarten K, Gorboulev V, and Koepsell H. The plasma membrane-associated protein RS1 decreases transcription of the transporter SGLT1 in confluent LLC-PK 1 cells. J Biol Chem 276: 45330-45340, 2001.' r: c& `# S0 ?
0 J* q+ q5 x2 u$ E9 d/ f1 O/ ^8 i0 K% N9 ~! O1 _5 D0 }7 K
) D$ q% `- z9 |0 xLambotte S, Veyhl M, Köhler M, Morrison-Shetlar AI, Kinne RKH, Schmid M, and Koepsell H. The human gene of a protein that modifies Na - D -glucose co-transport. DNA Cell Biol 15: 769-777, 1996.; q5 l8 t$ D1 G6 \& f* S
9 X' s$ Q5 z/ d7 M* r
1 a! p0 C: {8 {! j) c1 k
! {8 F/ H! `- b4 o" u# OLuzio JP, Brake B, Banting G, Howell KE, Braghetta P, and Stanley KK. Identification, sequencing and expression of an integral membrane protein of the trans -Golgi network (TGN38). Biochem J 270: 97-102, 1990.6 E% ~: d6 C/ b" ~: j
& @5 a3 w) D& ^# `7 T' [* C& E# o& Z1 U1 u/ J6 C! g/ B
8 `/ f+ A& W" S, H5 d
Marsh BJ, Mastronarde DN, Buttle KF, Howell KE, and McIntosh JR. Organellar relationships in the Golgi region of the pancreatic beta cell line, HIT-T15, visualized by high resolution electron tomography. Proc Natl Acad Sci USA 98: 2399-2406, 2001.* a# s0 C, h6 | R* f( ]
1 r" x$ `* D9 `
+ f8 B; Z8 F& k; w; K# y$ W
7 L& A. f0 G/ T0 |Nielsen MS, Madsen P, Christensen EI, Nykjaer A, Gliemann J, Kasper D, Pohlmann R, and Petersen CM. The sortilin cytoplasmic tail conveys Golgi-endosome transport and binds the VHS domain of the GGA2 sorting protein. EMBO J 20: 2180-2190, 2001.# O+ s. a# f/ ~/ Q/ A6 P+ P
( G( n" |8 k# k- F) r! ]7 `% [
" ~1 |: i6 x0 I* B/ U
4 o/ _5 R/ h- dOsswald C, Baumgarten K, Stümpel F, Gorboulev V, Akimjanova M, Knobeloch K-P, Horak I, Kluge R, Joost H-G, and Koepsell H. Mice without the regulator gene Rsc1A1 exhibit increased Na - D -glucose cotransport in small intestine and develop obesity. Mol Cell Biol 25: 78-87, 2005.
6 S/ @( y) M0 B- d: T2 x# E1 d7 m8 T. m& u
% y1 x2 x; R5 ^0 v$ R5 O$ a
, L. b& O3 l- n/ U3 W! DPoppe R, Karbach U, Gambaryan S, Wiesinger H, Lutzenburg M, Kraemer M, Witte OW, and Koepsell H. Expression of the Na - D -glucose cotransporter SGLT1 in neurons. J Neurochem 69: 84-94, 1997.# h* I4 g& M5 d# s1 y9 J
+ E& M, _# |6 O6 D# k1 h9 g. s
/ j- F8 f4 B0 q3 Y: w( |9 w2 Q: j6 y1 m# A
Reinhardt J, Veyhl M, Wagner K, Gambaryan S, Dekel C, Akhoundova A, Korn T, and Koepsell H. Cloning and characterization of the transport modifier RS1 from rabbit which was previously assumed to be specific for Na - D -glucose cotransport. Biochim Biophys Acta 1417: 131-143, 1999.
) W% ~4 `3 k. m2 o$ K: m# h F% _& B4 p! d$ a
# J/ G, \3 x; G0 n1 @* b3 E/ r3 |0 w# j$ _: A
Rios RM and Bornens M. The Golgi apparatus at the cell centre. Curr Opin Cell Biol 15: 60-66, 2003.
& ?9 `8 x# Z( R6 t- O0 C- r
; h* P5 g, K, Q/ ~- P- A: i
8 ^% q( k3 o2 T9 ~7 r" C" L2 Z& z$ M1 z6 C! k
Rios RM, Sanchis A, Tassin AM, Fedriani C, and Bornens M. GMAP-210 recruits -tubulin complexes to cis -Golgi membranes and is required for Golgi ribbon formation. Cell 118: 323-335, 2004.' z# |* t' M& V
6 u2 |7 _" l/ h3 C& V8 j# X: |
6 @- n% C Y8 |9 `# V7 T$ j' e7 {: o' G
Roth MG. Snapshots of ARF1: implications for mechanisms of activation and inactivation. Cell 97: 149-152, 1999., F$ Z) b- d: \1 @6 h6 G0 o
+ K6 k* \3 U* w
$ F6 x1 @$ X! \9 O. E
: `3 r# g$ x; T% ^; ]8 tScott PM, Bilodeau PS, Zhdankina O, Winistorfer SC, Hauglund MJ, Allaman MM, Kearney WR, Robertson AD, Boman AL, and Piper RC. GGA proteins bind ubiquitin to facilitate sorting at the trans -Golgi network. Nat Cell Biol 6: 252-259, 2004.% a3 C6 l( F/ I+ u4 A
# y0 O* j7 m$ c
' ?) r9 u- b% p t& ^
: A1 z4 W4 ]6 M7 nShioda T, Ohta T, Isselbacher KJ, and Rhoads DB. Differentiation-dependent expression of the Na /glucose cotransporter (SGLT1) in LLC-PK1 cells: role of protein kinase C activation and ongoing transcription. Proc Natl Acad Sci USA 91: 11919-11923, 1994.
" O/ i. w' _, U3 ^: R4 J
X: f. p! z! d( x; }) V. W$ t* t9 ^- E1 q4 T" t, Y- t
# y" `" A: E- p
Silver PA. How proteins enter the nucleus. Cell 64: 489-497, 1991.
( v2 m& c. _3 d6 T% \8 H; z% }) S& l. z! ~4 D# Q$ E! Y8 s
! F3 ]" N* }) R" R- n5 M
1 b$ ?6 S o4 p" `* E: P" u; ^Takatsu H, Katoh Y, Shiba Y, and Nakayama K. Golgi-localizing, -adaptin ear homology domain, ADP-ribosylation factor-binding (GGA) proteins interact with acidic dileucine sequences within the cytoplasmic domains of sorting receptors through their Vps27p/Hrs/STAM (VHS) domains. J Biol Chem 276: 28541-28545, 2001.3 A* o; G; y$ E- u
( G' d' z0 |4 A: s7 C( q% n' I R6 G% z; H h
; D& P0 m* J+ G; C* p
Thyberg J and Moskalewski S. Role of microtubules in the organization of the Golgi complex. Exp Cell Res 246: 263-279, 1999.% G6 w7 s1 M7 f; p7 w+ v. S
& a* v$ d# ]1 O
- v' p4 R8 O7 U' `; N/ W
' r e5 n. h4 V' dValentin M, Kühlkamp T, Wagner K, Krohne G, Arndt P, Baumgarten K, Weber W-M, Segal A, Veyhl M, and Koepsell H. The transport modifier RS1 is localized at the inner side of the plasma membrane and changes membrane capacitance. Biochim Biophys Acta 1468: 367-380, 2000.
8 {( b$ W3 ?# R2 t3 _: m% g: d/ T4 E
V( X, Z4 c% K7 k3 J H
3 c/ X4 p1 c8 }1 ?% u' z7 f0 W: ]0 z+ W1 D
van Dam EM and Stoorvogel W. Dynamin-dependent transferrin receptor recycling by endosome-derived clathrin-coated vesicles. Mol Biol Cell 13: 169-182, 2002.
, ^6 U' b8 E) V8 y' d k5 b5 H6 H3 r4 e$ u9 p
2 U8 Z6 M' E |
5 J" w; m$ i7 X1 O7 f LVeyhl M, Keller T, Gorboulev V, Vernaleken A, and Koepsell H. RS1 ( RSC1A1 ) regulates the exocytotic pathway of Na - D -glucose cotransporter SGLT1. Am J Physiol Renal Physiol 291: F1213-F1223, 2006.3 R) ^7 }7 x8 X8 j) ^
; p$ {9 ` U$ p; n% }6 e
7 J8 Z! ~- X8 P9 r/ r8 O L g
+ l) U6 f' b2 b! X# A$ \2 t6 uVeyhl M, Spangenberg J, Püschel B, Poppe R, Dekel C, Fritzsch G, Haase W, and Koepsell H. Cloning of a membrane-associated protein which modifies activity and properties of the Na - D -glucose cotransporter. J Biol Chem 268: 25041-25053, 1993.
& J4 S5 w& b: M$ I- l& f# ^5 Q
% ~- r0 [3 P* I# |& b; b0 _: K& A# ]- o2 G
3 Q, `' i3 Q7 m0 yVeyhl M, Wagner CA, Gorboulev V, Schmitt BM, Lang F, and Koepsell H. Downregulation of the Na - D -glucose cotransporter SGLT1 by protein RS1 (RSC1A1) is dependent on dynamin and protein kinase C. J Membr Biol 196: 71-81, 2003.
' p, g. T4 j% t! Y+ G' ?$ r5 N- {. ~' B: g
3 f' D# f- A7 K' C A3 M8 j' a1 ?0 E5 P8 N/ q* r
Wang P, Palese P, and O?Neill RE. The NPI-1/NPI-3 (karyopherin ) binding site on the influenza a virus nucleoprotein NP is a nonconventional nuclear localization signal. J Virol 71: 1850-1856, 1997.
5 K3 p3 G9 T7 B% [
, z, Z: `, e5 O% o4 `1 x$ y
$ C- _+ q4 D8 ^5 k
. B! q0 U, B* fWolff T, Unterstab G, Heins G, Richt JA, and Kann M. Characterization of an unusual importin alpha binding motif in the borna disease virus p10 protein that directs nuclear import. J Biol Chem 277: 12151-12157, 2002.
; V$ [1 ?3 v( j; g
" J' m9 {. R$ m* b- I
8 N7 L5 M6 e' }0 ^% z
2 R; O# h y% w7 HZorzano A, Muñoz P, Camps M, Mora C, Testar X, and Palacin M. Insulin-induced redistribution of GLUT4 glucose carriers in the muscle fiber. In search of GLUT4 trafficking pathways. Diabetes 45, Suppl 1: S70-S81, 1996. |
|
发表于 2009-4-22 08:32
